As the title implies, I'm hoping for some guidance in interpreting and resolving the subject error message on our server. This is an older Dell Poweredge 700 server running Windows 2003 Server, with a Paradise FastTrak RAID controller. The box is slated for retirement early next year, but must be maintained until retirement/replacement occurs.
In particular, this server hosts a 400GB three-disk RAID 5 array, consisting of three 200GB drives. One disk in the array failed last week, which was replaced and the array was automatically rebuilt successfully. Now, only two days after the rebuild, there are error events indicating bad blocks at the same LBA but on two disks, including the new drive eg
Disk ST3200822AS (Ch 1 Master) error at LBA 0x4a5356, Block Count 1 Disk ST3200822AS (Ch 3 Master) error at LBA 0x4a5356, Block Count 80
These two messages were logged approximately four seconds apart.
The drive on Ch 3 was the replaced/rebuilt drive. What puzzles me is that bad block messages could be generated from the same LBA across two physical drives. Is this simply an artifact of the way the disk blocks in the array are being mapped by the controller? Or is it possible that rebuilding the 3rd drive caused bad block data from the first drive to be migrated in some way to the new drive, and the bad block message itself for that disk isn't entirely accurate?
Given the expectation that the reports of bad blocks on the first drive are entirely accurate, the desire is to replace the drive. However, since that drive is still functional, is it still possible simply to replace it and allow the controller to rebuild it just as we did with the drive that actually failed (went offline)?
Many thanks in advance for any assistance.